AxonQA

MCP Integration

Model Context Protocol

Connect AxonQA to AI assistants like Claude Desktop, VS Code extensions, and OpenAI Codex using the Model Context Protocol (MCP).

Quick Start

1. Get API Key

Go to Settings > API Keys and generate a new key for MCP access.

2. Configure Client

Add AxonQA to your AI assistant using the configuration below.

3. Start Using

Ask your AI to import Jira tickets, generate tests, run automation, and more.

Client Configuration

Connect AxonQA to Claude Desktop for seamless AI-powered QA automation.

Step-by-Step Setup

  1. 1
    Get your API key

    Go to Settings → API Keys → Generate New Key

  2. 2
    Find your config file
    • macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
    • Windows: %APPDATA%\Claude\claude_desktop_config.json
  3. 3
    Add the configuration below

    Replace YOUR_API_KEY with your actual key

  4. 4
    Restart Claude Desktop

    Fully quit and reopen the app

json
{
  "mcpServers": {
    "axonqa": {
      "url": "https://your-axonqa-url.com/api/mcp",
      "transport": "http",
      "headers": {
        "Authorization": "Bearer YOUR_API_KEY",
        "Content-Type": "application/json"
      }
    }
  }
}

Try it out!

After setup, try asking Claude:

  • "Import Jira ticket PROJ-123"
  • "Generate test cases for the login feature"
  • "List my automation frameworks"

Available Tools

import_jira

Import Jira tickets as test requirements

generate_tests

Generate test cases from requirements

run_tests

Execute test suites locally

run_tests_browserstack

Run tests on BrowserStack cloud

capture_dom

Capture DOM elements for locators

list_frameworks

List automation frameworks

generate_automation_code

Generate Playwright/Cypress/Selenium code

open_framework_in_ide

Open framework in VS Code, Cursor, etc.

get_heal_suggestions

Get locator healing suggestions

apply_heals

Apply locator fixes

sync_framework

Sync code between cloud and local

check_browserstack_status

Check BrowserStack connection

Server tools
Agent tools
Cloud tools (BrowserStack)

View all tools: /api/mcp

Example Prompts

Here are some things you can ask your AI assistant once connected:

🎫 Jira Integration

  • "Import Jira ticket PROJ-123"
  • "Import all tickets from sprint 5"
  • "Show me the requirements from PROJ-456"

🧪 Test Generation

  • "Generate test cases for the login feature"
  • "Create tests from ticket PROJ-123"
  • "Improve my existing test cases"

🎭 Automation Code

  • "Generate Playwright code for my tests"
  • "Create a Cypress test suite"
  • "Export tests as Selenium scripts"

▶️ Test Execution

  • "Run my Playwright tests"
  • "Execute the login test suite"
  • "Show me the results of the last run"

🩹 Self-Healing

  • "Show me broken locators"
  • "Get heal suggestions for failed tests"
  • "Apply the suggested fixes"

☁️ Cloud Testing

  • "Is BrowserStack connected?"
  • "Run tests on Chrome and Safari on Windows 11"
  • "Test on iPhone 15 Pro and Galaxy S24"
  • "What browsers can I test on BrowserStack?"

💻 IDE Integration

  • "Open SwagLabs framework in VS Code"
  • "Open my tests in Cursor"
  • "Generate Cypress code and open in VS Code"

🔄 Multi-Step Workflows

  • "Import PROJ-123, generate tests, create Playwright code, then open in VS Code"
  • "Generate Cypress tests using SwagLab locators and run on BrowserStack"
  • "Run tests on Chrome and Firefox, then fix any broken locators"

API Reference

JSON-RPC Methods

MethodDescription
initializeHandshake and capability negotiation
tools/listList all available tools
tools/callExecute a tool with arguments
tools/infoGet detailed info about a tool
pingHealth check with agent status

Example: Call a Tool

json
{
  "jsonrpc": "2.0",
  "id": 1,
  "method": "tools/call",
  "params": {
    "name": "import_jira",
    "arguments": {
      "ticketId": "PROJ-123"
    }
  }
}

Endpoints

/api/mcp

Main JSON-RPC endpoint

/api/mcp/stream

SSE streaming for long-running tools

/api/mcp/openapi.json

OpenAPI 3.1 specification

/mcp/*.example.json

Configuration templates

Rate Limits & Security

Rate Limits

  • Default: 100 requests/minute
  • Headers: X-RateLimit-Limit, X-RateLimit-Remaining
  • 429 response when exceeded with Retry-After header

Security

  • Bearer token authentication required
  • API keys can have expiration dates
  • Destructive tools require confirmation
  • All tool executions are audit logged

Need Help?

Check the API endpoint for live tool information or contact support.

View API InfoGet API Key